Monday, April 29, 2013


Karen Martin wrote an excellent post a while back on Mark Graban's Lean Blog, so there's really no need to write this post, but I'm going to do it anyway because it's fun in a really nerdy kind of way.

Study, not Check

Anyway, like Karen and Dr. Deming, I also agree that the third step of the Shewhart Cycle for Learning and Improvement should be 'Study' as opposed to 'Check.'  Let's play word association with these two words:

  • When I think of 'Study' I think:  learning, impartial, scientist, curious, absorbing, methodical
  • When I think of 'Check' I think:  just verifying, yep I got what I expected, let's move on quickly
Doing a check is better than the approach most people use, which consists of implementing blindly toward an end point that never seems to arrive or which disappoints when it finally does.  But, 'Check' does not encourage a scientific mentality our curious learner attitude the way 'Study' does.  Improvement activities in their finest form are examples of the scientific method being applied to everyday opportunities for improvement, and we so often fail to understand that.  Hence, the quick check on the way to finishing a PDCA cycle and no value given to the learning part of the Shewhart Cycle.

Act, not Adjust (maybe)

Where I might (not definitely but maybe) diverge from Karen's view is with my preference of 'Act' over 'Adjust' as the fourth step of the Shewhart Cycle.  After studying the results of an experiment, there are actions to be performed regardless of whether adjustments are needed or not.  In practice, this is a moot point because adjustments are always needed, but I like to send the message through the use of the word 'Act' that action is required regardless of the outcome we achieved.  Other than adjustment, what actions might be required?  Here's a short list:
  • Sharing:  regardless of what we learned or what outcome we achieved, we should share our learning broadly across the organization i.e. yokoten.
  • Standardizing:  even if a change is not perfect (and will require adjustment), we might benefit from implementing the change into our work through the use of standard work, job instruction training, etc.
Semantics Matter

Again, I can see why 'Adjust' might be better and I'm not averse to it the way I am with 'Check.'  And yes, in the end it's just semantics, but when we're trying to establish a True North for how we go about learning and improving (which I hope all the lean coaches out there are doing on a daily basis), semantics matter.  A lot!